Cellular extracts from

HEK293T cells overexpressing EGL-18:HIS bind to the cat-1 CRM. Supershift of the band with HIS antibody, but not with GFP antibody, indicates that the binding involves EGL-18 protein. Moreover, point mutation of GATA site abolishes cat-1 sequence binding by the cellular extract. These experiments were fully performed by Dr. Miren Maicas.

bryo (Finney & Ruvkun 1990). We wanted to check if we could reproduce these results using the in- tegrated fosmid otIs337 strain (Zhang et al. 2014), so we could next proceed to analyse its expres- sion in different mutant backgrounds. With this configuration, the onset of expression and tissue specificity of unc-86 gene product can be easi- ly assessed. Expression was first detected in the HSN at embryonic comma stage (approximate- ly 7 hours post-fertilisation), and was maintained in adult worms → Figure 3.2.14-A and B. We next

crossed the otIs337 reporter with a tph-1 transcrip- tional reporter strain that carried a red fluorescent protein (vsIs97) and confirmed that unc-86 is also expressed in the NSM but not in the ADF neuron, al- though no quantification was carried out.

Basson and Horvitz first reported sem-4::lacZ fu- sion transgene expression in the HSN (Basson & Horvitz 1996). In the present work, in order to eval- uate sem-4 expression, we first used the integrat- ed fosmid wgIs57 (Sarov et al. 2012). Expression was hard to detect in the HSN during embryonic stages but GFP was localised to the HSN in L1 larva and it was maintained throughout the life of the animals

→ Figure 3.2.14-A. However, as fluorescence in-

tensity in the HSN was very faint and worms showed a relatively high level of background GFP signal, we next analysed a translational fusion reporter strain,

kuIs34, that had been previously described to be ex-

pressed in the adult HSN (Grant et al. 2000). A small- er percentage of L1 worms showed GFP expression in the HSN but, already at L2 stage, GFP penetrance was comparable to those of the fosmid report- er and expression was also maintained during the rest of the life of the worms → Figure 3.2.14-B. We

concluded that both reporters could be indistinct- ly used to study the HSN neuron. Similarly to unc-

86, sem-4 is expressed in many other neurons and

cells: hypodermal 8, 9 and 10 cells, all rectal cells, DVC, VPCs, ventral nerve cord neurons, cells in the head and in the preanal ganglion (Grant et al. 2000;

Jarriault et al. 2008). After crossing this kuIs34 re- porter with the tph-1::DsRed reporter strain vsIs79, we saw that sem-4 is not expressed in any other serotonergic neuron (not in NSM nor in ADF). According to some reports, hlh-3 appears to be expressed in all neuronal precursors during em- bryogenesis (Krause et al. 1997). A different report describes that, using either a full-length transla- tional reporter or transcriptional fusion reporters, this expression is maintained in most neurons of the nerve ring ganglia upon hatching and that already at larval stage 1, expression is almost undetecta- ble except for the endodermal-like P cells (Doonan et al. 2008). Expression is maintained in the 53 re- sulting postmitotic motor neurons, including HSN, throughout larval development. In our work, we chose two fosmid reporter strains to assess HLH-3 expression in the HSN. One was the otEx4140 ex- trachromosomal transgene (Murgan et al. 2015) and the other one was wgIs650 (Sarov et al. 2006), both constructs corresponding to recombineered fos- mids. In our hands, expression was only detected in the HSN precursor cell, the HSN/PHB neuroblast, at 5 hpf. Both reporters show comparable expression pattern and timing. → Figure 3.2.14-A and B show

expression of otEx4140. Expression was rapidly lost as we did not manage to see hlh-3::yfp expres- sion in the HSN at the time unc-86::yfp (otIs337) in- itiates its expression in the cell (430 min after the first cleavage, approximately 7 hpf). We mounted in parallel one cell stage embryos of the two reporter strains, otIs337 and otEx4140, and 4 hours later we scored the total number and the position of fluores- cent cells in the embryonic tail. We scored at dif- ferent time points and up to 6 hours after mounting the embryos (7 hpf). We never saw YFP express- ing cells at the same location and at the same time, so we concluded that hlh-3 expression in HSN/PHB precursor cell must be turned off before the last postmitotic division takes place, coinciding with

154 155

hlh-3 acting as a proneural gene during neurogen-

esis. We did not assess expression of hlh-3 in NSM or ADF precursor neurons.

egl-18 reporters and EGL-18 anti-sera had been

previously shown to be strongly expressed in seam cells, neurons in the head, VPCs and hyp7 cell, amongst others (Koh & Rothman 2001; Koh et al. 2002), but never before in the HSN cell. To study egl-

18 expression we used the fosmid strain stIs11606

(Dr. Waterston laboratory). We observed fluores- cent expression in the HSN in all larval stages and in adult worms → Figure 3.2.14-A and B. egl-18 ex-

pression in embryos was broad, so we assessed ex- pression in the embryonic HSN crossing unc-86::yfp reporter into the egl-18::mCherry background. Both reporters co-localised in the HSN postmitotic cell at 7 hpf. Using the tph-1 reporter zdIs13, we deter- mined that egl-18 is also expressed in the NSM neu- ron, but not in the ADF in adult worms.

For the two other TF members, however, there was no available fosmid. To study egl-46 expression we injected N2 worms with a transcriptional re- porter that covered all the intergenic region (4,477 bp) (kindly provided by Dr. Pocock) to create the

vlcEx324 extrachromosomic reporter line. Although DsRed expression was rather variable from scor-

ing to scoring, we detected expression in the HSN in all larval stages and in adult worms in 50-60% of the cases → Figure 3.2.14-A and B. We did not as-

sess expression in the embryo as it had been al- ready described that egl-46 expression in the HSN starts at 1.5 fold stage (Wu et al. 2001). As with the previous TFs, egl-46 is expressed in additional neu- rons in the head, ventral nerve cord and tail (Wu et al. 2001). We also determined that it is expressed in the NSM neuron, but not in the ADF neuron.

Finally, to examine ast-1 expression we first ana- lysed the fusion reporter line hdIs42 that contains the entire coding sequence of ast-1 (Schmid et al. 2006). YFP was observed in the nuclei of approxi- mately 40 neurons in the worm, but not in the HSN

neuron. As a way to test if the gene product of the

hdIs42 array was functional and rescued the ast-1

HSN phenotype, we crossed it with ast-1(hd92) le- thal mutants. The array was not able to rescue

ast-1 lethality, indicating that it missed relevant

information for its proper expression. As we be- lieved that ast-1 must be expressed in the HSN, in the same way as the rest of its partners, we decid- ed to generate CRISPR-Cas9 mediated GFP protein knock-in, tagging AST-1 protein at the C-terminus. More than ten integrated lines were recovered and all were undoubtedly expressed in the adult HSN

→ Figure 3.2.14-A and B. We chose one line to

study the expression and temporal pattern of the gene (ast-1(vlc19[ast-1::gfp])). Interestingly, AST- 1::GFP signal was first detected in the HSN at late L3 larval stage and was increasingly up-regulat- ed until the adult stage, when almost a 100% of the worms showed GFP in the cell. Moreover, it was ex- clusively detected in neuronal nuclei, in contrast to what was previously described (Schmid et al. 2006). Apart from HSN, we scored 30-32 neurons in the head, two HSNs, two PDEs and two ALN neurons in the tail plus another four neurons in the tail. We did not observe expression in the NSM or ADF seroton- ergic neurons.

In summary, the six members of the HSN regula- tory code are expressed in the HSN but the onset expression varies among the different TFs. unc-

86, sem-4, egl-18 and egl-46 are expressed from

early embryo or L1 stage and during the whole life of the worm, while ast-1 specifically acti- vates its expression in the cell in the transition be- tween L3 and L4 stages. The five of them would be co-expressed at the larval stage 4, when HSN differentiates and starts expressing most ter- minal features. In contrast, hlh-3 is only detect- ed in the HSN at embryonic stages, prior to the onset of HSN terminal differentiation. As afore- mentioned, bHLH TFs have a proneural role during neurogenesis and contribute to the specification

of progenitor-cell identity (Bertrand et al. 2002). This could suggest that hlh-3 acts earlier in the de- velopmental history of the HSN neuron. However, HSN sister neuron is not affected in hlh-3 mutant background → Figure 3.2.7. There are examples

in C. elegans where bHLH proneural genes, such as lin-32 and hlh-2, have also later roles in terminal neuron differentiation (Portman & Emmons 2000).

HSN terminal differentiation involves parallel pathways

To further explore how the HSN TF code regulates HSN terminal fate we aimed to assess if they show cross-regulation. We analysed TF expression of the previously analysed reporters in the different mu- tant backgrounds of young adult worms. We select- ed this age for analysis because the HSN neuron already should have become mature. For hlh-3 re- porter analysis, as it is only expressed in the em- bryo, expression was assessed in 5 hours old embryos. → Figure 3.2.15 shows how individual

factors affect the expression of the different TF re- porters. Although statistics between wild type and mutant values were calculated using raw data, we have represented in the graphs mutant values rela- tive to wild type values for an easier interpretation.

Moving on to the analysis, the ast-1(ot417) mutation does not affect the expression of any other HSN code member. The same happens with egl-

18(ok290) and egl-46(sy628) mutants, with the

exception of egl-46 mutants that show a small but significant decrease in ast-1 expression

→ Figure 3.2.15-A, → Annex 3.2.6. This implies

that AST-1, EGL-18 and EGL-46 TFs are not required for the expression of the rest of the code. Note that for egl-46 expression analysis in egl-18(ok290) mu- tants, kuIs34 strain could not be used due to an in- compatibility in chromosome location. Instead, the

kuIs35 transgene was used, which corresponds to

an independent integration event of the same con- struct as kuIs34 (Grant et al. 2000). Although we did not characterise kuIs35 in the same depth as

kuIs34, both transgenes seemed to show identical

expression patterns.

In contrast to the previous cases, UNC-86 seems to be required for the expression of several mem- bers of the HSN TF code. We observed a severe phenotype in unc-86(n846) mutants upon ast-1 ex- pression, sem-4 expression and egl-46 expression.

sem-4(n1971) and hlh-3(tm1688) mutants, in turn,

also seem to affect ast-1 expression. hlh-3(tm1688) mutants show an additional mild phenotype over

egl-46 expression → Figure 3.2.15-A. Our results

EGL-18 tph-1 bas-1 cat-1 UNC-86 UNC-86 AST-1 AST-1 UNC-86 Figure 3.2.13

Summary of EMSAs: UNC-86, AST-1 and EGL-18 bind to serotonin pathway gene CRMs in electrophoretic mobility assays

UNC-86 binds to tph-1, cat-1 and bas-1. AST-1 binds to

bas-1 and cat-1, but not to tph-1, at least with the probe

and under the conditions tested. EGL-18 binds to cat-1, but not to either tph-1 probe tested.

157 L4 Adult Embr yo 0 50 100

HSN regulatory code expression

L1 L2 L3 % HSN unc-86::yfp sem-4::gfp hlh-3::yfp egl-46::DsRed ast-1::gfp egl-18::mCherry Figure 3.2.14 Expression of the HSN regulatory code in the HSN neuron